K5: Testing RAW Continuous Shooting

Started Oct 20, 2010 | Discussions
steelski Senior Member • Posts: 2,555
Re: DNG or PEF?

GordonBGood wrote:

k2park wrote:

Dale108 ? Have you tried with PEF? PEF must be compressed one, could be better.
Could you confirm it?

I doubt that PEF is any different than DNG as since the K-7 (K-7, K-x, K-r, and K-5), DNG's have been compressed by almost the same algorithm as PEF's.

Regards, GordonBGood

Whats the size of an uncompressed 14bit DNG

GordonBGood Veteran Member • Posts: 6,308
Re: DNG or PEF?

steelski wrote:

GordonBGood wrote:

k2park wrote:

Dale108 ? Have you tried with PEF? PEF must be compressed one, could be better.
Could you confirm it?

I doubt that PEF is any different than DNG as since the K-7 (K-7, K-x, K-r, and K-5), DNG's have been compressed by almost the same algorithm as PEF's.

Whats the size of an uncompressed 14bit DNG

Uncompressed DNG's are a bit strange in the the Lossless JPEG compression standard states that for all bit depth above 12, byte packing will not be used and therefore it requires two bytes per raw photosite. As the K-5 records 4992 by 3284 raw photosites, multiply the product of these by two to get the size of the uncompressed raw data, then add a few KBytes of meta data, the small thumbnail, and the extra up to approximately 2.7 MBytes for the highly compressed JPEG preview embedded in the raw for a total of about 35 MBytes maximum for the K-5.

Regards, GordonBGood

richardplondon
richardplondon Forum Pro • Posts: 10,762
Re: PEF slower?

Gerry Winterbourne wrote:

k2park wrote:

Dale108 ? Have you tried with PEF? PEF must be compressed one, could be better.

On my K10 I started using PEFs because they were compressed and DNGs weren't, so I could get more on a card. Then I got fed up with the separate XMP files and switched to DNG. I noticed that write time was quicker: I don't know for certain but I've always put that down to the extra time to compress PEFs.

The balance of advantage here will vary depending on the speed of the card - compressing takes time but means there is less to write. Shooting PEF or DNG may not be relevant to whether XMP are used in postprocessing, since it is easy to convert to compressed DNG at import time (I also have a K10d, and use Lightroom). But separate XMPs simply don't bother me, personally, atm.

Once I upgrade to a camera that can do it, camera-compressed native DNG may well be my choice thereafter - although, it must be said, regular sync backup is a lot easier when transferring just a small XMP for each changed image, rather than the whole DNG package.

RP

 richardplondon's gear list:richardplondon's gear list
Pentax K-5 Pentax smc DA 21mm F3.2 AL Limited Pentax smc DA 35mm F2.8 Macro Limited Pentax smc DA 70mm F2.4 AL Limited Pentax smc DA 18-55mm F3.5-5.6 AL WR +6 more
Dale108
OP Dale108 Veteran Member • Posts: 7,358
Re: DNG or PEF?

For PEFs, I was able to get 8 RAW files but that was at slow shutter speeds in indoors light at ISO 1600 so not comparable to my first test. I will try and test under good lighting conditions that will simulate wildlife shooting.

Dale
--
http://www.pbase.com/abundant108

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=272176&subSubSection=1787360&language=EN

 Dale108's gear list:Dale108's gear list
Sony RX10 III Olympus TG-5 Nikon Coolpix P1000 Pentax K-1 Panasonic Lumix DC-G9 +1 more
SanMat Senior Member • Posts: 2,861
Re: K5: Testing RAW Continuous Shooting Using 16 MB Sandsisk Extreme

Thanks for posting all the info Dale.

Will be very interested to hear how it goes with your birds in flight shots,

Pete

Dale108 wrote:

At ISI 320 indoors, I am able to shoot 6 RAW files in a burst and then there is about a 1 sec delay and I can shoot another 3 shots. This card allows 30MB/s. Possibly a faster card will help. I put the K5 into MF mode so focus was not an issue as this will slow down a burst. Can't wait to try on some birds in flight to see how it does.

Dale
--
http://www.pbase.com/abundant108

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=272176&subSubSection=1787360&language=EN

MightyMike Forum Pro • Posts: 38,598
Re: 128mb buffer in my estimation...

if the cost is anything like the RAM we stick in our computers then the cost per camera would be minimal, think we can get gigs of RAM relatively cheap, so why would 128mb of nRAM be expensive?

anyhow GordonbGood always brings up good points but for now i believe its simply the limitation of card write speed. The fact that a 30MB/s card doesn't guarantee 30MB/s all the time but at most will guarantee 10MB/s, this shows us that the card is the weak link and an estimated average of 15MB/s should be expected for the best cards.

Buffer sizes won't improve much at all on any brands until card write speeds get up to a guaranteed 120MB/s, that's going to take some time!
--
Mike from Canada

'I like to think so far outside the box that it would require a telephoto lens just to see the box!' ~ 'My Quote :)'

http://www.airliners.net/search/photo.search?sort_order=views%20DESC&first_this_page=0&page_limit=180&&emailsearch=mighty_mike88%40hotmail.com&thumbnails=

Zvonimir Tosic
Zvonimir Tosic Senior Member • Posts: 2,775
Did you check lens correction?

Dale108 wrote:

One of the only issues I had with the K5 specs is the size of the RAW buffer during continuous shooting compared to the K5. I picked up a K5 this afternoon and in one quick test was only able to shoot 5-6 RAW files before the buffer filled up.

With Lens Corrections on in your menu, that's what you can expect.

I was shooting 14 RAWs on a K-7 in sequence just now without lens correction. Same result you have by default when using older manual focusing lenses -- 14 RAWs (Lens Correction menu item is then disabled).

But, when turning lens corrections on, I got just 5 RAWs.
All images taken at ISO 1600 btw.

Ta.

dgaies Regular Member • Posts: 290
Re: Did you check lens correction?

It's not because of the lens corrections (it would be nice if that was why).

With lens correction off, the number of raw files that can be recorded in one burst goes from 8 at iso 100 down to 4 above iso 6400. FYI, the file size (DNG) goes from about 18 MB to 32 MB as you increase ISO (exact sizes will vary depending on scene)

Zvonimir Tosic wrote:

Dale108 wrote:

One of the only issues I had with the K5 specs is the size of the RAW buffer during continuous shooting compared to the K5. I picked up a K5 this afternoon and in one quick test was only able to shoot 5-6 RAW files before the buffer filled up.

With Lens Corrections on in your menu, that's what you can expect.

I was shooting 14 RAWs on a K-7 in sequence just now without lens correction. Same result you have by default when using older manual focusing lenses -- 14 RAWs (Lens Correction menu item is then disabled).

But, when turning lens corrections on, I got just 5 RAWs.
All images taken at ISO 1600 btw.

Ta.

Zvonimir Tosic
Zvonimir Tosic Senior Member • Posts: 2,775
Thank you

dgaies wrote:

It's not because of the lens corrections (it would be nice if that was why).

With lens correction off, the number of raw files that can be recorded in one burst goes from 8 at iso 100 down to 4 above iso 6400. FYI, the file size (DNG) goes from about 18 MB to 32 MB as you increase ISO (exact sizes will vary depending on scene)

This is good to know about K5.
Thank you very much for the info!

Dale108
OP Dale108 Veteran Member • Posts: 7,358
Re: Did you check lens correction?
 Dale108's gear list:Dale108's gear list
Sony RX10 III Olympus TG-5 Nikon Coolpix P1000 Pentax K-1 Panasonic Lumix DC-G9 +1 more
GordonBGood Veteran Member • Posts: 6,308
Re: 128mb buffer in my estimation...

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per camera would be minimal, think we can get gigs of RAM relatively cheap, so why would 128mb of nRAM be expensive?

Mike, I'm pretty sure that the K-7/K-5 contain at least 256 MBytes of RAM if they store raw files in compressed form, and likely much more that that if they are stored in non-compressed form as per my last post. Knowing which is the case should be fairly simple, as for minimum file sizes as with the lens cap on, raw buffer image depth should be greatly increased above normal images if images are stored in compressed format, where the buffer depth will stay the same if they are stored in uncompressed format. I have asked dgaies to do a test, which he indicates he will, likely within the next little while

anyhow GordonbGood always brings up good points but for now i believe its simply the limitation of card write speed. The fact that a 30MB/s card doesn't guarantee 30MB/s all the time but at most will guarantee 10MB/s, this shows us that the card is the weak link and an estimated average of 15MB/s should be expected for the best cards.

Yes, the card may be somewhat the bottleneck, although it is likely also the camera's interface to those cards. Notice that the Canon T2i, a lesser camera, can output its 18 MP 14-bit raw files at a minimum of almost 25 MBytes at about one a second, tested with these same Sandisk Extreme III 30 MBytes per second SDHC cards. Also, the K-7 had a full buffer raw frame rate of about 1.4 frames per second as tested by DPR with typical raw file sizes of about 15 MBytes at low ISO's for an output rate averaging over 20 MBytes per second and including time for image acquisition and processing.

Buffer sizes won't improve much at all on any brands until card write speeds get up to a guaranteed 120MB/s, that's going to take some time!

Mike, I'm afraid you're wrong here. Buffer depths aren't affected by write speeds at all; card write speeds only affect the time it takes to clear the buffer and the buffer full rate. If the K-5 had double the buffer and could typically buffer about 16 raw images (assuming there is a bug that is causing the buffer size to halve for high ISO images), everyone would be pretty much satisfied, even if the buffer full rate was about 1 frame per second and it took almost 20 seconds to fully clear the buffer at current card write speeds.

Regards, GordonBGood

MightyMike Forum Pro • Posts: 38,598
Re: 128mb buffer in my estimation...

GordonBGood wrote:

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per camera would be minimal, think we can get gigs of RAM relatively cheap, so why would 128mb of nRAM be expensive?

Mike, I'm pretty sure that the K-7/K-5 contain at least 256 MBytes of RAM if they store raw files in compressed form, and likely much more that that if they are stored in non-compressed form as per my last post. Knowing which is the case should be fairly simple, as for minimum file sizes as with the lens cap on, raw buffer image depth should be greatly increased above normal images if images are stored in compressed format, where the buffer depth will stay the same if they are stored in uncompressed format. I have asked dgaies to do a test, which he indicates he will, likely within the next little while

The RAW data may be compressed right at the processing engine and therefore never be uncompressed in the buffer. the data comes into the processing engine, its processed, compressed and spit out as a compressed RAW file to the buffer. Now of course i could be wrong.

If this helps... on the K-7 i got 14 frames, a slight pause then 1 frame with the lens cap on at low ISO, same with the lens cap off... in the same way i got 12+1 at ISO6400 lens cap off and 13 even with the lens cap on.

anyhow GordonbGood always brings up good points but for now i believe its simply the limitation of card write speed. The fact that a 30MB/s card doesn't guarantee 30MB/s all the time but at most will guarantee 10MB/s, this shows us that the card is the weak link and an estimated average of 15MB/s should be expected for the best cards.

Yes, the card may be somewhat the bottleneck, although it is likely also the camera's interface to those cards. Notice that the Canon T2i, a lesser camera, can output its 18 MP 14-bit raw files at a minimum of almost 25 MBytes at about one a second, tested with these same Sandisk Extreme III 30 MBytes per second SDHC cards. Also, the K-7 had a full buffer raw frame rate of about 1.4 frames per second as tested by DPR with typical raw file sizes of about 15 MBytes at low ISO's for an output rate averaging over 20 MBytes per second and including time for image acquisition and processing.

Buffer sizes won't improve much at all on any brands until card write speeds get up to a guaranteed 120MB/s, that's going to take some time!

Mike, I'm afraid you're wrong here. Buffer depths aren't affected by write speeds at all; card write speeds only affect the time it takes to clear the buffer and the buffer full rate. If the K-5 had double the buffer and could typically buffer about 16 raw images (assuming there is a bug that is causing the buffer size to halve for high ISO images), everyone would be pretty much satisfied, even if the buffer full rate was about 1 frame per second and it took almost 20 seconds to fully clear the buffer at current card write speeds.

I disagree to a point... one of the biggest complaints about the Sigma DP1 and similar cameras is how slow it is, now this is an extreme case as described below

  • Shot to Shot time RAW as 4.5 seconds

  • RAW Burst at 3.4fps for 3 frames with a buffer clear time of a massive 13.5 seconds

If Pentax considers an approx. buffer clear time of 10 seconds to be acceptable for the majority then 20 seconds would be a point to complain about IMO. Sure you may have burst off for 2 seconds instead of 1 and have double the shots. But in a fast paced action even 20 seconds will seem a lot more like an eternity then 10 seconds. I do understand that one can fire off smaller burst while waiting for the buffer but you can't review the initial burst set until the buffer is empty. Next up lets match the K-7s performance of approx 3 seconds and 15 shots. well 15 shots is only a 2 second burst but if it were desired to have a 3 second burst then we'd be looking at 21-22 shots, now the clear time is 30 seconds.

So what do people desire, burst off more shots or for longer periods of time or some overlap, and then how long do they think is acceptable before they can do it again. I think Pentax is right at about 10 seconds and if this holds true then we won't be seeing any buffer size increase until the write speed (camera and card) improves.

Regards, GordonBGood

-- hide signature --

Mike from Canada

'I like to think so far outside the box that it would require a telephoto lens just to see the box!' ~ 'My Quote :)'

http://www.airliners.net/search/photo.search?sort_order=views%20DESC&first_this_page=0&page_limit=180&&emailsearch=mighty_mike88%40hotmail.com&thumbnails=

GordonBGood Veteran Member • Posts: 6,308
Re: 128mb buffer in my estimation...

MightyMike wrote:

GordonBGood wrote:

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per camera would be minimal, think we can get gigs of RAM relatively cheap, so why would 128mb of nRAM be expensive?

...for minimum file sizes as with the lens cap on, raw buffer image depth should be greatly increased above normal images if images are stored in compressed format, where the buffer depth will stay the same if they are stored in uncompressed format.

The RAW data may be compressed right at the processing engine and therefore never be uncompressed in the buffer. the data comes into the processing engine, its processed, compressed and spit out as a compressed RAW file to the buffer. Now of course i could be wrong.

Yes, it is technically possible to compress the raw data using the image processing engine as the data stream is read from the sensor, but then the camera would have to be decompressed to generate the JPEG thumbnails and previews as well as the accompanying JPEG if Raw+ mode is used. I think it unlikely, as decompression is actually harder as in more time consuming to do than compression. However, the raw sensor data could be streamed (Direct Memory Access - DMA?) to an area of memory and another process/thread could be compressing that into the raw buffer along with converting that same raw data to the various JPEG's required.

If this helps... on the K-7 i got 14 frames, a slight pause then 1 frame with the lens cap on at low ISO, same with the lens cap off... in the same way i got 12+1 at ISO6400 lens cap off and 13 even with the lens cap on.

So therefore the K-7 is not dependent on compressed file size; I would assume that the K-5 will be the same but dgaies will soon let us know. This would indicate that uncompressed raw data is being saved in the buffer and its hard to see where all the memory got used up as the raw data should just be 32.8 divided by 30 = less than 10% bigger.

anyhow GordonbGood always brings up good points but for now i believe its simply the limitation of card write speed. The fact that a 30MB/s card doesn't guarantee 30MB/s all the time but at most will guarantee 10MB/s, this shows us that the card is the weak link and an estimated average of 15MB/s should be expected for the best cards.

Yes, the card may be somewhat the bottleneck, although it is likely also the camera's interface to those cards. Notice that the Canon T2i, a lesser camera, can output its 18 MP 14-bit raw files at a minimum of almost 25 MBytes at about one a second, tested with these same Sandisk Extreme III 30 MBytes per second SDHC cards. Also, the K-7 had a full buffer raw frame rate of about 1.4 frames per second as tested by DPR with typical raw file sizes of about 15 MBytes at low ISO's for an output rate averaging over 20 MBytes per second and including time for image acquisition and processing.

Buffer sizes won't improve much at all on any brands until card write speeds get up to a guaranteed 120MB/s, that's going to take some time!

Mike, I'm afraid you're wrong here. Buffer depths aren't affected by write speeds at all; card write speeds only affect the time it takes to clear the buffer and the buffer full rate. If the K-5 had double the buffer and could typically buffer about 16 raw images (assuming there is a bug that is causing the buffer size to halve for high ISO images), everyone would be pretty much satisfied, even if the buffer full rate was about 1 frame per second and it took almost 20 seconds to fully clear the buffer at current card write speeds.

I disagree to a point... one of the biggest complaints about the Sigma DP1 and similar cameras is how slow it is, now this is an extreme case as described below

  • Shot to Shot time RAW as 4.5 seconds

  • RAW Burst at 3.4fps for 3 frames with a buffer clear time of a massive 13.5 seconds

The Sigma cameras aren't really a common case for "raw image formats" as their sensor is different and their imaging processor is slow. Notice that it takes almost as long to clear about the same three images from the buffer after a JPEG burst, indicating that most of the time is spent processing the images and not writing the data to flash card.

If Pentax considers an approx. buffer clear time of 10 seconds to be acceptable for the majority then 20 seconds would be a point to complain about IMO. Sure you may have burst off for 2 seconds instead of 1 and have double the shots. But in a fast paced action even 20 seconds will seem a lot more like an eternity then 10 seconds. I do understand that one can fire off smaller burst while waiting for the buffer but you can't review the initial burst set until the buffer is empty. Next up lets match the K-7s performance of approx 3 seconds and 15 shots. well 15 shots is only a 2 second burst but if it were desired to have a 3 second burst then we'd be looking at 21-22 shots, now the clear time is 30 seconds.

So what do people desire, burst off more shots or for longer periods of time or some overlap, and then how long do they think is acceptable before they can do it again. I think Pentax is right at about 10 seconds and if this holds true then we won't be seeing any buffer size increase until the write speed (camera and card) improves.

You are likely right about not increasing the buffer size, so therefore better use has to be made of the buffer that is there. If we had an optional 12-bit raw output mode just as the K-7 has, we should than have almost all of it, with about 13 shots to fill the buffer, 1.2 frames per second or a little better when the buffer is full, and still about 10 seconds to clear the buffer. Remember that you don't have to wait the full 10 seconds for the buffer to fully clear with Pentax cameras to be able to start shooting some again, just wait until there is room for another few pictures.

Regards, GordonBGood

Freestyler Regular Member • Posts: 170
Re: 128mb buffer in my estimation...

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per > camera would be minimal, think we can get gigs of RAM relatively cheap, so why > would 128mb of nRAM be expensive?

It's not the same thing, RAM in computers is technically very slow, you are talking similar to cache on CPU's which only have maybe 1-12Mb cache on die (current Intel CPU's), it's fast, small and works on very little power and it also has take up very little room on the chips that they are going on, depending on how many banks of what size ram @ variable nm die sizes and who manufactuers them (usually Toshiba and Samsung)

GordonBGood Veteran Member • Posts: 6,308
I'm pretty sure the raw buffer is at least 256 MBytes...

I'm pretty sure the raw buffer is at least 256 MBytes, but the total RAM memory is likely much more than that as there would be required large buffers for other purposes. For instance, there is very likely a 48.6 MByte buffer to hold the developed RGB imag for JPEG processing, etc. This is based on the following:

GordonBGood wrote:

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per camera would be minimal, think we can get gigs of RAM relatively cheap, so why would 128mb of nRAM be expensive?

Mike, I'm pretty sure that the K-7/K-5 contain at least 256 MBytes of RAM if they store raw files in compressed form, and likely much more that that if they are stored in non-compressed form as per my last post. Knowing which is the case should be fairly simple, as for minimum file sizes as with the lens cap on, raw buffer image depth should be greatly increased above normal images if images are stored in compressed format, where the buffer depth will stay the same if they are stored in uncompressed format. I have asked dgaies to do a test, which he indicates he will, likely within the next little while

I got the results of the requested testing from dgaies by PM, and it seems that the buffer depth stays the same for any raw output file size and any ISO up to 6400, where all variations cause about a halving of the raw buffer depth. To me, this indicates that Pentax are saving the raw data as uncompressed data at 32.8 Mbytes per buffered image, and then very likely developing the RGB image to output as reduced size thumbnail and highly compressed embedded full size JPEG image as well as compressing the raw data as part of the output conversion routines.

Further, I can surmise that, using the same net amount of raw buffer memory, for the 12-bit raw output cameras Pentax "byte packed" the 12 bit sensor output on acquisition so that each photosite storage only took a byte and a half, whereas in preparation for output of the 14-bit raw files they now use two bytes per photosite. It may well also be faster to acquire the data in this way. Thus, the space required per image is about four thirds bigger than the increase in raw MP.

Therefore, for a K-7 that had a raw buffer depth of about 15 but which took about 3 seconds to fill that buffer by which time about three images would have already been written out, had a raw buffer of about 12 images time 1.5 times the raw MP of 15 or about 270 MBytes. Now if we take that same space and apply the K-5's high speed continuous shooting to it, we get that there will only be about a second before the raw buffer files, which means that only about one image will have already been written out. 270 MBytes at two times 16.4 raw MP is 30.8 MBytes means that only about eight raw images would fit into it. Thus one gets about eight to an absolute maximum of nine for smaller file sizes and fast write times, just as is seen.

As to the halving of the raw buffer depth for ISO's of 6400 and higher, I now imagine that a new type of fancy high ISO Noise Reduction (NR) is used for these higher ISO's that requires an additional results output buffer that is the same size as that of the raw data itself. What this NR does or its presence may be very difficult to prove mathematically, as it may be some "sneaky" type of additional Nikon/Sony like type of chroma NR. It looks like it may be intentional and therefore not a bug, and that K-5 users may just have to live with this characteristic in order to get the observed high ISO performance.

Regards, GordonBGood

GordonBGood Veteran Member • Posts: 6,308
Re: 128mb buffer in my estimation...

Freestyler wrote:

MightyMike wrote:

if the cost is anything like the RAM we stick in our computers then the cost per > camera would be minimal, think we can get gigs of RAM relatively cheap, so why > would 128mb of nRAM be expensive?

It's not the same thing, RAM in computers is technically very slow, you are talking similar to cache on CPU's which only have maybe 1-12Mb cache on die (current Intel CPU's), it's fast, small and works on very little power and it also has take up very little room on the chips that they are going on, depending on how many banks of what size ram @ variable nm die sizes and who manufactuers them (usually Toshiba and Samsung)

Digital cameras don't really need very fast RAM (in modern terms), as the fastest thing they do is read the raw data from the sensor. For the K-5 that's about 16.4 million words of data at least 7 times per second for a clock rate of only about 114.8 million cycles per second. That is nothing for any DDR type of RAM. The typical camera micro-controller type of CPU usually only clocks at 100's of MHz, so that doesn't really need fast memory either. And most memory operations work as a scan from one end of the image array to the other, so the latency in setting up rows and columns also isn't really a factor.

As has already been discussed with Mike, the reason that the K-5 has this amount of memory, whatever it should be, is that the K-7 had it, but also with twice the memory there would be twice the buffer clearing time of about 20 seconds, which might well be unacceptable.

Regards, GordonBGood

Johan Josefsson Senior Member • Posts: 1,158
Re: K5: Testing RAW Continuous Shooting

Hi,

It seems I missed this discussion, but according to the DPreview specs on the preview it should be almost the same.

K-5
Continuous Hi: 7 fps, up to 40 Best JPEG frames, 15 images RAW*

K-7
Continuous Hi: 5.2 fps, up to 40 Best JPEG frames, 15 images RAW, 14 DNG*

But then pentaximaging.com specs says
Continuous FPS - Continuous Hi: 7.0 FPS (22 JPG, 8 RAW)

So, the DPreview preview specs were wrong all along?

-- hide signature --

Regards,
Johan - Swe

http://joj.smugmug.com

dgaies Regular Member • Posts: 290
Re: K5: Testing RAW Continuous Shooting

So, the DPreview preview specs were wrong all along?

It seems there was some inaccurate information floating around out there regarding the K5 (in regards to the buffer) prior to its release.

Dale108
OP Dale108 Veteran Member • Posts: 7,358
Re: K5: Testing RAW Continuous Shooting

Hi Johan:

I can only report my experience which does not support the DPreview specs; I wish they were right.

Dale
--
http://www.pbase.com/abundant108

http://www.pentaxphotogallery.com/home#section=ARTIST&subSection=272176&subSubSection=1787360&language=EN

 Dale108's gear list:Dale108's gear list
Sony RX10 III Olympus TG-5 Nikon Coolpix P1000 Pentax K-1 Panasonic Lumix DC-G9 +1 more
dlacoutu Contributing Member • Posts: 811
Re: PEF slower?

richardplondon wrote:

The balance of advantage here will vary depending on the speed of the card - compressing takes time but means there is less to write. Shooting PEF or DNG may not be relevant to whether XMP are used in postprocessing, since it is easy to convert to compressed DNG at import time (I also have a K10d, and use Lightroom). But separate XMPs simply don't bother me, personally, atm.

Problem is that the Adobe Raw Converter strips information from the PEFs while converting them to DNGs...

The black pixel stripe, for instance, is removed, but it can be helpful (even essential on a K20!).

From memory, it does not so when going from a K10 DNG to a compressed DNG.

 dlacoutu's gear list:dlacoutu's gear list
Pentax K-5 Samsung NX100 Samsung NX200 Pentax smc DA 18-55mm F3.5-5.6 AL WR Pentax smc FA 50mm F1.4 +6 more
Keyboard shortcuts:
FForum MMy threads