[Radiohead]

The DPR review of the Alpha 700 is up

Some interest for Nikonians/D300 buyers here as there is a lot of comparison with the D300. Most noticeable is that the Sony never gives you a RAW file without some form of NR.

[springtide]

>> Most noticeable is that the Sony never gives you a RAW file without some form of NR.


Yep, that is true for ISO1600 shots and above. If you look at the block diagram for the processing, you'll see that NR is actually performed on the CMOS sensor prior to any further NR processing.

The NR was a lot stronger in firmware v1 (particularly with Jpegs) and it appears that the v2 firmware backed off this NR but it is assumed in the processing section and not on the sensor. Firmware v2 was released after a few reviews had commented on this heavy handed NR.

So as stated above, for High ISO (1600+), there is no "Off" setting, only "Low".

There are now more rumours of a v3 firmware for the A700. Whether this is a fact or just a rumour - only time will tell.

What does this mean in real life? Well, to me it doesn't really make any difference at all. If I take high ISO photos, then having on-chip NR performed seems to me the best place to try and do it, even if you loose a little bit of sharpness. I guess if I had invested in 3rd party NR applications then this would be more of an issue.

The point which I thought was interesting, is that ISO200 is actually the lowest ISO setting on the A700. And "ISO100" is implemented in the same way at the D300.

Update: Firmware v3 has been released. No additional NR changes though!

[Pirate!!]

http://support.sony-europe.com/dime/...sp?l=en&f=A700

I don't have the A700, and there's speculation what v3 actually does other than what it says on the tin . . but it's a free upgrade for those who have it.

FIXES:

1. When pictures are taken rapidly, the access lamp may stay lit and the system may become inoperable.

2. When turning off the camera when the "MR" mode is selected, some settings registered in the MR mode can be reflected in other shooting modes

[springtide]

Originally Posted by Pirate!!

http://support.sony-europe.com/dime/...sp?l=pl&f=A700

I don't have the A700, and there's speculation what v3 actually does . . but it's a free upgrade for those who have it.

Improvements over Firmware Ver.2:-
- When pictures are taken rapidly, the access lamp may stay lit and the system may become inoperable.
- When turning off the camera when the "MR" mode is selected, some settings registered in the MR mode can be reflected in other shooting modes.

So just bug fixes by the look at it! I must admit, I'd not seen these issues myself, but a few people on another forum had.

Sony should hold their head in shame for releasing a "buggy" camera!
I guess it is good news that they seem to care enough to bother to release new firmware.

[springtide]

Radiohead, just noticed that you also posted the same statement in the "Sony Alpha A700 (Body) - Sub £800.00" thread!!!

I assume you are a D300 fan

There are one of two questions that I have with the DPR review now that I've had a good read...

Within the review, they state based on the IQ seem when using RAW of the A700 and D300 that the camera must be using the same CMOS sensor. This is not unusual, since I think the D2Xs also used a Sony sensor and this rumour has been around for a while (I hav read somewhere that the D3 sensor was done in house).

The review then goes on to state, that because the initial NR is being done on the sensor of the A700, that the RAW files will always have NR applied.
(btw, There is the other argument that says the RAW file is the output of the sensor, hence if NR is applied on chip, then the RAW file does not have NR applied, since the RAW is the output of the sensor and not the CMOS cells. This argument does has it's merit, but I think in a court of law the case neither side would win!)

Anyway, getting back to the original unanswered question, if the NR is being applied on the sensor and it cannot be turned off (can you see where were are going with this...), then if the D300 shares the same sensor... hmmm, doesn't that mean that the on-chip NR is being applied to the D300 RAW files? (that is, if they are the same sensor).

This is not a ...."lets have a pop at the D300", but more if a "do we think it would be possible to turn off the NR that is being seen on the A700 at high ISO settings?"

Sorry if there are spelling mistalkes in thies post, typed very quicklyu as I'm cookling supper£!!

[Radiohead]

Originally Posted by springtide

I assume you are a D300 fan

My sig should tell you what I shoot with...

[springtide]

Originally Posted by Radiohead

My sig should tell you what I shoot with...

Obviously spotted your sig, but I think I saw you post somewhere that you'd had a play with a D300 ?

[Yandros]

Originally Posted by springtide

Anyway, getting back to the original unanswered question, if the NR is being applied on the sensor and it cannot be turned off (can you see where were are going with this...), then if the D300 shares the same sensor... hmmm, doesn't that mean that the on-chip NR is being applied to the D300 RAW files? (that is, if they are the same sensor).

No. Check out page 18...

http://www.dpreview.com/reviews/sonydslra700/page18.asp


"Below we have a comparison of in-camera JPEG to RAW converted using Adobe Camera RAW with its Noise Reduction options set to zero (effectively 'no noise reduction'); RAW NONR. As you can see the Nikon D300's RAW images are just that, warts and all (but at least this means you can run your own noise reduction on them if you wish and you are getting the 'digital negative'). The A700 on the other hand clearly already has some luminance noise reduction applied to its RAW data which leaves the photographer with no option other than to simply accept Sony's decisions on noise reduction. "

The pics certainly confirm that too.

[springtide]

@Yandros: Yes, thats what I'm talking about..

There are two NR systems in the A700, one on the CMOS sensor itself, the other within the image processor.

So the issue which is being implied/stated on this forum (not sure if this was implied or not), Dpreview and other KM/Sony forums is that Sony will never be able to turn off the NR on the RAW files in firmware due to the fact that this is being performed "on-chip" (the CMOS sensor).

Sony has already addressed NR "reduction" once already when they released firmware v2. They basically turned down the amount of NR being applied at high ISO's (see the Sony link), but this was mostly related to Jpegs and not RAW, and there is still no "off" setting. Can they do anymore in firmware is the question?

But this is not the case with the D300, which is rumoured to have the same CMOS sensor (If you read the text, DPreview were pretty convinced it is the same sensor). If this is the case and the D300 doesn't have the same heavy handed NR as the A700, then what I'm saying/speculating is that either a) the sensors are different, b) this heavy handed NR is being done on the image processor c) the on-chip NR is programmable (there could be more opts)

So in most of the cases above, it should be possible to turn off this NR in firmware.

So the basic question is, "is it possible for Sony to fully turn off this NR to give similar results at high ISO as the D300, by tweaking the firmware?" Obviously most of the time you'll want the NR enabled when using high ISO, but some people want the option of using 3rd party applications for NR as part of PP.

I'm not sure if I explained that very well. Make sense?

[Yandros]

Ahh, sorry *penny drops*. I see what you mean now. I've no idea if it's on-chip or firmware. I hope for Sony's sake that it's firmware, coz the serious PPers will riot . I certainly mess around with selective NR on 800+ ISO images, so I'd hate it for a start.