Reading mode: Light Dark

About the E-300 SLR

Go to last post
T

tsk1979

Senior Member
Messages
1,186
Reaction score
283
Location
Delhi, Noida, IN
The review has generated quite a furore ;). Though i am not a photographer i work in the VLSI field and i can answer some of the questions which came up.

One was Phil wondering wether the quirks can be fixed by newer firmware. The answer is yes. The Kodak CCD based cameras are mostly running on the OMAP platform(TI) and similar chipsets. These are extremely customizable. Infact if you do not worry about warranty you can actually write your own firmware.

In the world of digital media players etc., this is done quite frequently. I guess photographers who will actually spot the differnce are not realy into software and stuff most of the time.

The other thing in the review is the shadow noise. This is again an algorigthm problem. The gain settings in the DSP program are not proper here. This leads to a high signal to noise ration. Looking at the review and also going through the specs it seems that that camera is actually much better than it actually seems to be.

The so called flaws are software fixable. My take on this is that if Olympus had not been under the pressure of Xmas, and spent maybe a couple of months on the software extra, this camera would have got HR rating even. I am hoping that within 6 months or so they come up with a firmware upgrade.

Before you wonder what magic can a piece of software do, take a bite on this. EOS 300D hacked firmware makes it at par with the EOS 10D functionally and performance wise.
--
Tanveer
 
Jack Riley said:
What did it do for the sensor/noise characteristics.
Click to expand...
Am not too sure as the article was sketchy. All it said was that it bumped the EOS300D to EOS10D level. Actually you can get pretty good noise characteristics if you are willing to sacrifice some speed. The noise reduction algorithms which are used professionaly use lots of CPU time. So if you go for a better algorithm you can see speed dropping to about half. Currently burst speed is limited by card write speed, but with the best of noise reduction algos they could be the limiting factor.

--
Tanveer
 
tsk1979 said:
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.
Click to expand...
That's an extremely interesting insight. I hadn't considered hacking a camera before. I wonder which operating system they are using. TI lists a number of choices for their platform. Perhaps Linux? How about an opensource camera? Gimp in a camera?

Gotta run and take a cold shower....

Cheers, Morton
 
Thanks for an informative post. We need more of those here at the moment... ;-)

I agree with you that we may hope for better algos in the next FW version, but what I find surprising is that Oly provides similar performance in their RAW conversion software that runs on PC/Mac. Where there's more processing power and more time...
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.

In the world of digital media players etc., this is done quite
frequently. I guess photographers who will actually spot the
differnce are not realy into software and stuff most of the time.

The other thing in the review is the shadow noise. This is again an
algorigthm problem. The gain settings in the DSP program are not
proper here. This leads to a high signal to noise ration. Looking
at the review and also going through the specs it seems that that
camera is actually much better than it actually seems to be.

The so called flaws are software fixable. My take on this is that
if Olympus had not been under the pressure of Xmas, and spent maybe
a couple of months on the software extra, this camera would have
got HR rating even. I am hoping that within 6 months or so they
come up with a firmware upgrade.

Before you wonder what magic can a piece of software do, take a
bite on this. EOS 300D hacked firmware makes it at par with the EOS
10D functionally and performance wise.
--
Tanveer
Click to expand...
 
It is a relief having read about that very objective view on the kodak sensor. I'm neither & photo pro nor a program pro but sheer synapse tell me that the flaws are just a wares thing. Oly should just come-up with an upgrade (firm & soft) sooner to avoid competitive losses especially with the detailed review Phil wrote.

By the way ****, an off-note... I saw your gallery on your photo site, the photos are GREAT!!! If it's ok with you, I would love to some copies via email. Thanks cheerdude!
**** Rijken said:
I agree with you that we may hope for better algos in the next FW
version, but what I find surprising is that Oly provides similar
performance in their RAW conversion software that runs on PC/Mac.
Where there's more processing power and more time...
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.

In the world of digital media players etc., this is done quite
frequently. I guess photographers who will actually spot the
differnce are not realy into software and stuff most of the time.

The other thing in the review is the shadow noise. This is again an
algorigthm problem. The gain settings in the DSP program are not
proper here. This leads to a high signal to noise ration. Looking
at the review and also going through the specs it seems that that
camera is actually much better than it actually seems to be.

The so called flaws are software fixable. My take on this is that
if Olympus had not been under the pressure of Xmas, and spent maybe
a couple of months on the software extra, this camera would have
got HR rating even. I am hoping that within 6 months or so they
come up with a firmware upgrade.

Before you wonder what magic can a piece of software do, take a
bite on this. EOS 300D hacked firmware makes it at par with the EOS
10D functionally and performance wise.
--
Tanveer
Click to expand...
Click to expand...
 
Is Olympus here? Do they read these forums? Are they aware of what needs/should be done? Who do we email? Is there a way to go about requesting/suggesting firmware upgrades?

CG
tsk1979 said:
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.
Click to expand...
--
Chris Anderson
Nature, Cave, and Celestial Imagery
http://www.darklightimagery.net
C-8080 Gallery
http://www.pbase.com/darklightimagery/root
 
They're over 5 MB each, so I can attach two per email message. If you give me some filenames, I can email originals. I found a yahoo email address on your site. Can I use that? Can you receive large attachments there? You may want to check how much room there is in your mailbox... Let me know...

Cheers -- ****
Chris Alcala MD said:
By the way ****, an off-note... I saw your gallery on your photo
site, the photos are GREAT!!! If it's ok with you, I would love to
some copies via email. Thanks cheerdude!
**** Rijken said:
I agree with you that we may hope for better algos in the next FW
version, but what I find surprising is that Oly provides similar
performance in their RAW conversion software that runs on PC/Mac.
Where there's more processing power and more time...
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.

In the world of digital media players etc., this is done quite
frequently. I guess photographers who will actually spot the
differnce are not realy into software and stuff most of the time.

The other thing in the review is the shadow noise. This is again an
algorigthm problem. The gain settings in the DSP program are not
proper here. This leads to a high signal to noise ration. Looking
at the review and also going through the specs it seems that that
camera is actually much better than it actually seems to be.

The so called flaws are software fixable. My take on this is that
if Olympus had not been under the pressure of Xmas, and spent maybe
a couple of months on the software extra, this camera would have
got HR rating even. I am hoping that within 6 months or so they
come up with a firmware upgrade.

Before you wonder what magic can a piece of software do, take a
bite on this. EOS 300D hacked firmware makes it at par with the EOS
10D functionally and performance wise.
--
Tanveer
Click to expand...
Click to expand...
Click to expand...
 
I'm sure they are aware of what their camera can and can't do. I think time constraints is the most likely answer here.
CG Anderson said:
CG
tsk1979 said:
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.
Click to expand...
--
Chris Anderson
Nature, Cave, and Celestial Imagery
http://www.darklightimagery.net
C-8080 Gallery
http://www.pbase.com/darklightimagery/root
Click to expand...
 
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
Click to expand...
Tanveer,

Do you know anybody who can hack the E-300 firmware? Firmware copy should be available once Olympus releases the new version. :)
 
I thought there might have been some hacks for the e-10, coming out of Slovenia or Montenegro. They do some wonderful things with the Yugo over there!

Barry
 
Don't get too optimistic about what can be done by hackers with the firmware. The 300D hacks are based on the fact that the 300D firmware was really the 10D firmware that Canon had artificially crippled to remove features. The hackers merely worked their way round the Canon blocks to reenable what was already there. The hacks did nothing to the image processing, only provided access to features such as one shot focusing, FEC etc.
tsk1979 said:
Jack Riley said:
What did it do for the sensor/noise characteristics.
Click to expand...
Am not too sure as the article was sketchy. All it said was that it
bumped the EOS300D to EOS10D level. Actually you can get pretty
good noise characteristics if you are willing to sacrifice some
speed. The noise reduction algorithms which are used professionaly
use lots of CPU time. So if you go for a better algorithm you can
see speed dropping to about half. Currently burst speed is limited
by card write speed, but with the best of noise reduction algos
they could be the limiting factor.

--
Tanveer
Click to expand...
 
tsk1979 said:
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.
Click to expand...
From the Olympus website:

"The exclusive TruePic TURBO image processor delivers fast shooting and operating speeds while helping to ensure smooth, clear image quality."

Are you saying that Olympus has in fact not developed their own custom image processor LSI (a la Canon Digic) and that they are simply using "off the shelf" parts?

You should realize that writing firmware from scratch is virtually impossible without detailed knowledge of the hardware.
tsk1979 said:
The other thing in the review is the shadow noise. This is again an
algorigthm problem. The gain settings in the DSP program are not
proper here. This leads to a high signal to noise ration.
Click to expand...
Are you saying that Olympus simply needs to change a couple of parameters (gain settings) to fix the shadow noise problem?
By the way, a high signal to noise ratio is a good thing.
tsk1979 said:
The so called flaws are software fixable.
Click to expand...
How can you be sure without knowing all of the details of the "exclusive TruePic TURBO image processor"?
tsk1979 said:
Before you wonder what magic can a piece of software do, take a bite
on this. EOS 300D hacked firmware makes it at par with the EOS 10D
functionally and performance wise.
Click to expand...
It is obvious that a firmware change can enable features that were disabled by firmware, no?

--
'There is no conversation more boring than the one where everybody agrees.'
 
I love that chip! So easy to program, and fast and low power. Actually I think the ARM-based chips have the best MIPS/Watt ratio of any chip.
Thank you for that tidbit.
J.
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.
Click to expand...
--
http://www.4-3system.com/
http://jonr.light.is/
http://getfirefox.com/
 
Barry Stewart said:
I thought there might have been some hacks for the e-10, coming out
of Slovenia or Montenegro. They do some wonderful things with the
Yugo over there!

Barry
Click to expand...
That would require far more work for them than the E-1 or E-300, which already are designed to update via the internet. The E-10 is not, so they'd first have to figure out how to write the firmware first before cracking its code and tweaking it. In another thread it was mentioned that the update is done by Olympus repair centers through the CF-slot -- if this were true, it's completely possible to jury-rig a firmware updater for the E-10 and E-20.

Also, the Canon Rebel 300 hack was an unlocker, as mentioned above. Image processing was not affected. To update the image processing of the CCD, you would not only need a firmware hacker but one who has access to and knows how to program image processing algorithms. That's a tall order for any programmer!

--
http://linus.grumblemutterspit.org/photos.php
 
butachan said:
How can you be sure without knowing all of the details of the
"exclusive TruePic TURBO image processor"?
Click to expand...
Actually this is all marketing Jargon.
TURBO, TruPIc all is marketing speak

Most common companies base the systems on Either ARM/MIPS for the housekeeping stuff and the DSP from TI mostly.
butachan said:
It is obvious that a firmware change can enable features that were
disabled by firmware, no?
Click to expand...
Yes it is true. As for the signal processing, Any decent DSP wouls suffice. But we are talking about a camera here. Would you want a camera that gave excellent images but took 5 seconds to take a shot? This is the reason a compromise would be needed.

For example in burst mode you can currently take a pic in 1/2 second. You could develop a beter algo which reduces your burst mode to 1 pic per second. This will not be acceptable.

And changing firmware where signal processig is involved is not a trivial task. Companies take about 6-7 months just testing the firmware after the processor is out. Not to mention another 6 months of testing before the silicon is out.

And that SnR was a typo. :)

And i know whati am talking about. I work in chip design(TI) and our company develops the OMAP platform for lots of comapines. The full list is on the TI website, but Kodak is there definately.
--
Tanveer
 
bstarling said:
Don't get too optimistic about what can be done by hackers with the
firmware. The 300D hacks are based on the fact that the 300D
firmware was really the 10D firmware that Canon had artificially
crippled to remove features. The hackers merely worked their way
round the Canon blocks to reenable what was already there. The
hacks did nothing to the image processing, only provided access to
features such as one shot focusing, FEC etc.
Click to expand...
Thats true too. But what i am suggesting is not impossible if you are willing to sacrifice performance. But the fact is that there are very few camera hackers in the world who are well versed with Signal Processiging
--
Tanveer
 
FFS said:
Tanveer,

Do you know anybody who can hack the E-300 firmware? Firmware copy
should be available once Olympus releases the new version. :)
Click to expand...
Hey FFS, the story was on slashdot.org too, you can probably search the site. Its doable but your warranty goes kaput.
As for firmware copy i think you can read it of the camera too.
--
Tanveer
 
TSK1979,

I feel you are in the software field rather than in VLSI field. The DSP is programmable does not mean the problems can easily fixed. The problem with Bill Gates' recent presentation is one example. I suspect the DSP is not fast enough so better algorithm cannot be implemented. On a PC, the raw processing is not limited by real time and the quality is better. On the other hand, Canon seems to be making good progress with the DSP performance (DIGIC II). That helps to get 8 fps on 1DII and 5 fps on 20D. You would hope the upcoming E-3 can match at least 20D.

The noise is determined by the sensor noise and signal processing. It is not that easy to improve without sacrifying detail.

The 300D hack has nothing to do with performance. It is just some features already in the firmware but disabled. The hack allows user to access it, mainly FEC, MLU and ISO3200.

Mapleaf
tsk1979 said:
The review has generated quite a furore ;). Though i am not a
photographer i work in the VLSI field and i can answer some of the
questions which came up.
One was Phil wondering wether the quirks can be fixed by newer
firmware. The answer is yes. The Kodak CCD based cameras are mostly
running on the OMAP platform(TI) and similar chipsets. These are
extremely customizable. Infact if you do not worry about warranty
you can actually write your own firmware.

In the world of digital media players etc., this is done quite
frequently. I guess photographers who will actually spot the
differnce are not realy into software and stuff most of the time.

The other thing in the review is the shadow noise. This is again an
algorigthm problem. The gain settings in the DSP program are not
proper here. This leads to a high signal to noise ration. Looking
at the review and also going through the specs it seems that that
camera is actually much better than it actually seems to be.

The so called flaws are software fixable. My take on this is that
if Olympus had not been under the pressure of Xmas, and spent maybe
a couple of months on the software extra, this camera would have
got HR rating even. I am hoping that within 6 months or so they
come up with a firmware upgrade.

Before you wonder what magic can a piece of software do, take a
bite on this. EOS 300D hacked firmware makes it at par with the EOS
10D functionally and performance wise.
--
Tanveer
Click to expand...
 
Great! Thanks ****! I still have around 200mb left in my yahoo email address and still 1 gb in my walla.com address bearing the same identity. Yahoo permits upto 10mb per msg (outbound & inbound). As to choices, I have nothing in particular as I like of your shots & style. Any will do as as you prefer. Hey, am I getting excited or what? LOLOL!!!

Cheers "cheer-up dude"! - Chris
**** Rijken said:
They're over 5 MB each, so I can attach two per email message. If
you give me some filenames, I can email originals. I found a yahoo
email address on your site. Can I use that? Can you receive large
attachments there? You may want to check how much room there is in
your mailbox... Let me know...

Cheers -- ****
Click to expand...
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top