Buffer Capacity Estimates

Yes the R3 has lower pixel density, so what? That has nothing to do with noise levels.
But it magnifies the noise thus making it more visible.
Prove that the R3 noise, of which has less noise than the R7, has more visible noise than the R7, in the real world. Show me real world examples.

EDIT: Use the link below to compare the R7 to the R3 at 6400 ISO by clicking on the image for an unaltered view. Look just above the crayons and just below the pillows on the left hand side. Even magnifying the R3 image to make it bigger than the R7 image, the noise doesn't look nearly as bad as the R7. Here is the link:

 
Last edited:
Right but it isn’t larger enough to account for the extra MP and frame rate so users are back to the bad old days of worrying about buffer.
You got 3 seconds of full speed RAW shooting with the 7DII. With the R7, using CRAW (and there's simply no reason not to)
Actually I often dig into my R5/R6 (RAW) files very deeply, and do notice that there is enough of a difference to warrant continued use of my Full RAW files (vs CRAW). I did the tests early on. So I do have to remain aware of my buffer as I shoot (which fortunately doesn’t become an issue often with these bodies).

With my (future) R7 I’ve already planned my usage cases to account for the (full RAW) buffer size, and should be OK.

You guys (shooting smaller files) needn’t worry as much! :-)

R2
 
Why go with the R7 at all than? You should get the R3 than or wait for the R1. That seems like the upgrade for you. I have the 7D MKII and do consider the R7 an upgrade, spec-wise as well as the reviews I've seen. I haven't even used my 7D MKII for quite a while, opting for my 5D MKIV instead most of the time. YMMV
You pay in other ways, though, besides the price.

The R3 has much lower pixel density, and it has a little bit more visible noise than the R7 in focal-length-limited and "equivalent" DOF situations.

The visible noise ranking above base ISO when total subject light is the same, as far as I can see, is:

R6 and R7 - the best from Canon

R5 and R3 - the second best group

All others at least a little noisier than the R5 and R3 (except the M6-II and the 90D, which are closer to the R6 and R7).

Even a 1.4x TC puts less pixels-on-subject than the R7; a 2x puts slightly more, but now your AF is dealing with 1/4 the light.
I find it very difficult to believe that a FF sensor camera "more visible noise" in any circumstance. Comparing my 7D MKII and my 5D MKIV is no comparison in noise levels, The 5D MKIV destroy the crop sensor 7D MKII every time.
Every time you magnify it less with the same exposure? Every time you get shallower DOF, too, because of the optics chosen?

The fact is, a 1.6x crop from the 5D4 has the same pretty much the same noise as the 7D2, and the 7D2 is a bit better in incandescent light. The fact that you never noticed this doesn't make it any less true. Same for when you use 2.5x the ISO on the 5D4 for the same needed DOF.

There's no generic FF noise benefit for a given amount of sensor area, or with a given amount of total light. You need more light at base ISO or with shallower DOF at any ISO, to get less noise with a FF camera.
Yes the R3 has lower pixel density, so what? That has nothing to do with noise levels.
It doesn't, but it affects the camera choice, especially among the people who were waiting for an R7 as a high-density camera rather than an inexpensive R camera (the RP is already even less expensive).
The FF sensor will display less noise than a cropped sensor every time. BTW, if I could afford the R3, I'd also buy the 600mm F4 or 500mm F4 lens and get be performance and noise making the comparison a moot point.
The full FF sensor will have less noise at the same exposure, but there are many reasons why, in practice, you won't fill the frame or will have less exposure, depending on situation and optics.
 
In a thread that was recently locked, the topic came up of how many files fit in the R7's buffer, which is rather small for a camera with a "7" in its name, implying that it is the mirror-less "7D2 successor".
People have been saying this, but it seems they have simply forgotten what the buffer on the 7DII is. The 7DII RAW buffer is 31 shots. A lot more in JPEG, obviously. I never, in six years of shooting my 7DII, hit the buffer limit. That's mostly because I did most of my sports shooting in JPEG, to fit more on the card and for ease and speed of editing. But even in JPEG I pretty much never shot bursts of more than 30. But the R7 buffer (at 15fps--50% faster than the 7DII's top speed) is reported at about 51 shots in RAW, over 100 in CRAW, and lots and lots in JPEG. Even the DPR hands-on with R7 claimed the buffer was smaller than the 7DII. But it simply isn't. Whether the R7 buffer counts as small is obviously a matter of perspective. But it's bigger than the 7DII. And I don't recall hearing many (in fact any) complaints about the buffer on the 7DII.
So, today I tried with another camera, the R5, to see how camera settings affect estimated buffer size (and card capacity as well), and it was about as simple as can be: nothing but ISO setting, and cRAW vs RAW affected the estimate (well, crop mode on the R5, but the R7 doesn't have that). This means that the estimate can be way, way off.

HTP and changing the shutter mode made no difference to the estimate, and "Auto ISO" assumes the worst case of the highest ISO.

There are things that we know change the compressed file sizes for a given scene; mainly noise and utilization of the tones or bits. You can't change the DR of your scene, but you can change how it is recorded. Under-exposure leaves lots of contiguous, unused most significant bits, and lower analog gains also result in less noise, which makes them easier to compress. HTP (Highlight Tone Priority) does both of these things; ISO 800 with HTP is the same RAW data as ISO 400 with a stop of under-exposure, two steps towards a smaller file. I would suggest to anyone who is going to do actual measurements of how many files the buffer can take before filling, that they explore HTP and maybe even a little bit more underxposure.

I used HTP for years until I realized that the R5's headroom was better than other cameras I used before, and the WYSIWYG "exposure simulation" made HTP less necessary, but if my future R7 gives me buffer grief, I may go back to using HTP.
With the 7d mk2 in jpeg mode you pretty much couldn’t run out of buffer. I never shoot RAW, partly because of the buffer limitations. When you’re used to having a nice big buffer it’s a little annoying to upgrade to a small one. Not saying you can’t get the job done with the R7 but it sure is nice to not have to worry about buffer.
But my point was that the R7 buffer is actually larger than the 7DII buffer. Why do people keep saying it's smaller? It isn't. It's actually considerably larger. If the upgrade you're talking about is from the 7DII to the R7, you are, in fact, upgrading from a "nice big buffer" to a considerably bigger one. See my post above to John where I give the actual numbers. It's not even close. The R7 is a much much bigger buffer, both in RAW and JPEG, than the 7DII. And when you add the ability to shoot CRAW, which the 7DII doesn't have, the increase in buffer size is even larger. This is one of the few things that we can know without trying out the camera. Just look at the specs.
And my point is that with the R7 we’re back to having to worry about buffer when that seemed to be a thing of the past with the 7d mk2. With the 7d mk2, the camera was always ready for the next burst, the R7, not so much.
Given the actual specs, that seems to be just plain false. The R7 has a bigger buffer than the 7DII in all respects, including how long you can shoot at 15fps. As to whether the R7 is always ready for the next burst, I have no idea. I don't have mine yet, and neither do you, so perhaps you should hold off? My 7DII was always ready for the next burst, but I never maxed it out. Given that, I expect I'll never max out my R7 either, which, as I have said repeatedly, has a much bigger buffer than the 7DII. I find it interesting that you first complain that the R7 has a smaller buffer than the 7DII, then, when I point out the easily available fact that it doesn't, complain that there isn't progress, then, when I point out that that it false too, fall back on a claim about it not being 'always ready for the next burst', when you simply have no way of knowing that. If you maxed out your 7DII, it wouldn't have been always ready for the next burst, because there would be times when it was clearing the buffer. If you didn't max out your 7DII, you won't maxing out your R7 at 50% faster burst rates, and definitely not at lower speeds. Why not simply admit that you are just wrong about the buffer of the R7 in comparison with the 7DII? You might still want the R7 to have a larger buffer, but stop making factually incorrect comparisons with the 7DII.
Well I’m basing that off of reports from actual users. I certainly hope you’re correct and that I can find out first-hand. I’m not talking about RAW by the way. As I mentioned earlier, I don’t use RAW so for me what counts is the jpeg buffer, which for the 7d mk2 was pretty much unlimited unless you were using a slow card.



in the video linked in this post, 7sec buffer with cRAW. That seems to validate your JPEG needs will have minimal issues.
 
In a thread that was recently locked, the topic came up of how many files fit in the R7's buffer, which is rather small for a camera with a "7" in its name, implying that it is the mirror-less "7D2 successor".
People have been saying this, but it seems they have simply forgotten what the buffer on the 7DII is. The 7DII RAW buffer is 31 shots. A lot more in JPEG, obviously. I never, in six years of shooting my 7DII, hit the buffer limit. That's mostly because I did most of my sports shooting in JPEG, to fit more on the card and for ease and speed of editing. But even in JPEG I pretty much never shot bursts of more than 30. But the R7 buffer (at 15fps--50% faster than the 7DII's top speed) is reported at about 51 shots in RAW, over 100 in CRAW, and lots and lots in JPEG. Even the DPR hands-on with R7 claimed the buffer was smaller than the 7DII. But it simply isn't. Whether the R7 buffer counts as small is obviously a matter of perspective. But it's bigger than the 7DII. And I don't recall hearing many (in fact any) complaints about the buffer on the 7DII.
So, today I tried with another camera, the R5, to see how camera settings affect estimated buffer size (and card capacity as well), and it was about as simple as can be: nothing but ISO setting, and cRAW vs RAW affected the estimate (well, crop mode on the R5, but the R7 doesn't have that). This means that the estimate can be way, way off.

HTP and changing the shutter mode made no difference to the estimate, and "Auto ISO" assumes the worst case of the highest ISO.

There are things that we know change the compressed file sizes for a given scene; mainly noise and utilization of the tones or bits. You can't change the DR of your scene, but you can change how it is recorded. Under-exposure leaves lots of contiguous, unused most significant bits, and lower analog gains also result in less noise, which makes them easier to compress. HTP (Highlight Tone Priority) does both of these things; ISO 800 with HTP is the same RAW data as ISO 400 with a stop of under-exposure, two steps towards a smaller file. I would suggest to anyone who is going to do actual measurements of how many files the buffer can take before filling, that they explore HTP and maybe even a little bit more underxposure.

I used HTP for years until I realized that the R5's headroom was better than other cameras I used before, and the WYSIWYG "exposure simulation" made HTP less necessary, but if my future R7 gives me buffer grief, I may go back to using HTP.
With the 7d mk2 in jpeg mode you pretty much couldn’t run out of buffer. I never shoot RAW, partly because of the buffer limitations. When you’re used to having a nice big buffer it’s a little annoying to upgrade to a small one. Not saying you can’t get the job done with the R7 but it sure is nice to not have to worry about buffer.
But my point was that the R7 buffer is actually larger than the 7DII buffer. Why do people keep saying it's smaller? It isn't. It's actually considerably larger. If the upgrade you're talking about is from the 7DII to the R7, you are, in fact, upgrading from a "nice big buffer" to a considerably bigger one. See my post above to John where I give the actual numbers. It's not even close. The R7 is a much much bigger buffer, both in RAW and JPEG, than the 7DII. And when you add the ability to shoot CRAW, which the 7DII doesn't have, the increase in buffer size is even larger. This is one of the few things that we can know without trying out the camera. Just look at the specs.
And my point is that with the R7 we’re back to having to worry about buffer when that seemed to be a thing of the past with the 7d mk2. With the 7d mk2, the camera was always ready for the next burst, the R7, not so much.
Given the actual specs, that seems to be just plain false. The R7 has a bigger buffer than the 7DII in all respects, including how long you can shoot at 15fps. As to whether the R7 is always ready for the next burst, I have no idea. I don't have mine yet, and neither do you, so perhaps you should hold off? My 7DII was always ready for the next burst, but I never maxed it out. Given that, I expect I'll never max out my R7 either, which, as I have said repeatedly, has a much bigger buffer than the 7DII. I find it interesting that you first complain that the R7 has a smaller buffer than the 7DII, then, when I point out the easily available fact that it doesn't, complain that there isn't progress, then, when I point out that that it false too, fall back on a claim about it not being 'always ready for the next burst', when you simply have no way of knowing that. If you maxed out your 7DII, it wouldn't have been always ready for the next burst, because there would be times when it was clearing the buffer. If you didn't max out your 7DII, you won't maxing out your R7 at 50% faster burst rates, and definitely not at lower speeds. Why not simply admit that you are just wrong about the buffer of the R7 in comparison with the 7DII? You might still want the R7 to have a larger buffer, but stop making factually incorrect comparisons with the 7DII.
Well I’m basing that off of reports from actual users. I certainly hope you’re correct and that I can find out first-hand. I’m not talking about RAW by the way. As I mentioned earlier, I don’t use RAW so for me what counts is the jpeg buffer, which for the 7d mk2 was pretty much unlimited unless you were using a slow card.
https://www.dpreview.com/forums/post/66325864

in the video linked in this post, 7sec buffer with cRAW. That seems to validate your JPEG needs will have minimal issues.
I’ve been told by people who I know and trust that have their R7’s already and shoot jpeg that they have had the “camera busy” pause while using the R7 for multiple short bursts. This isn’t something I experienced with the 7d mk2. Hopefully I’ll have my own R7 before 2023 to try for myself. This doesn’t make the R7 a useless or bad camera but it is a limitation to be aware of and has been confirmed by pretty much every reviewer. I don’t see any point in denying it or trying to dismiss it. I’m sorry if this is offensive to you.
 
They probably have DLO turned on HIGH. It really slows the write speed to the card down because of the extra processing. Even when shooting just raw with DLO on HIGH the camera write speed slows down the same amount. Could be a bug. DLO set to standard doesn't have this problem.
 
They probably have DLO turned on HIGH. It really slows the write speed to the card down because of the extra processing. Even when shooting just raw with DLO on HIGH the camera write speed slows down the same amount. Could be a bug. DLO set to standard doesn't have this problem.
Right. To avoid all this, I shoot RAW and turn all in-camera processing options OFF.

R2
 
They probably have DLO turned on HIGH. It really slows the write speed to the card down because of the extra processing. Even when shooting just raw with DLO on HIGH the camera write speed slows down the same amount. Could be a bug. DLO set to standard doesn't have this problem.
 
I went to take some (generic) photos of humpback whales from a boat yesterday....

At some stage two whales breached out of the water, I got 73 craw files in a single burst. I released the shutter because the whales had disappeared back in the water but didn't get any buffer issue and the camera was ready to take more shots.
Using EFSC 15fsp CRAW with a Sandisk fast card.
 
I went to take some (generic) photos of humpback whales from a boat yesterday....

At some stage two whales breached out of the water, I got 73 craw files in a single burst. I released the shutter because the whales had disappeared back in the water but didn't get any buffer issue and the camera was ready to take more shots.
Using EFSC 15fsp CRAW with a Sandisk fast card.
With C-RAW and 15 fps, I tested and got 146 images before the buffer was full.
 
I went to take some (generic) photos of humpback whales from a boat yesterday....

At some stage two whales breached out of the water, I got 73 craw files in a single burst. I released the shutter because the whales had disappeared back in the water but didn't get any buffer issue and the camera was ready to take more shots.
Using EFSC 15fsp CRAW with a Sandisk fast card.
With C-RAW and 15 fps, I tested and got 146 images before the buffer was full.
This confirms how awful and limited the R7 buffer is then :-)
 
I went to take some (generic) photos of humpback whales from a boat yesterday....

At some stage two whales breached out of the water, I got 73 craw files in a single burst. I released the shutter because the whales had disappeared back in the water but didn't get any buffer issue and the camera was ready to take more shots.
Using EFSC 15fsp CRAW with a Sandisk fast card.
With C-RAW and 15 fps, I tested and got 146 images before the buffer was full.
This confirms how awful and limited the R7 buffer is then :-)
That is never the debate here, as these shot counts are already known from the specs listed in the manual. The issue is someone will eventually come and say "this isn't enough for them" without context of what exact shooting use case it hampers or what other similar bodies do it better

Nothing wrong in wanting more, but I don't see how R7 won't do well in most situations with what it already has
 
I went to take some (generic) photos of humpback whales from a boat yesterday....

At some stage two whales breached out of the water, I got 73 craw files in a single burst. I released the shutter because the whales had disappeared back in the water but didn't get any buffer issue and the camera was ready to take more shots.
Using EFSC 15fsp CRAW with a Sandisk fast card.
With C-RAW and 15 fps, I tested and got 146 images before the buffer was full.
 

Keyboard shortcuts

Back
Top