Re: Media Testing for maclife.de (was: Re: Colorsync-users Digest, Vol 5, Issue 290)
Re: Media Testing for maclife.de (was: Re: Colorsync-users Digest, Vol 5, Issue 290)
- Subject: Re: Media Testing for maclife.de (was: Re: Colorsync-users Digest, Vol 5, Issue 290)
- From: "edmund ronald" <email@hidden>
- Date: Tue, 9 Sep 2008 17:21:20 +0200
Mark,
All of us on list use printers. We all know about canned printer
profiles - that's exactly how good Adobe's profiles are: They are
generic profiles, carefully made, and no better and no worse than any
other canned profile. Canned profiles are exactly as good as the
ability and the desire of the manufacturer to ensire product
consistency at that point.
Incidentally, Microsoft does not have this problem with canned
profiles, because they demand that camera manufacturers supply the
codecs for their Raw formats directly, and therefore the customer will
get an image that's decoded as well as the camera manufacturer can or
wants. Manufacturer codecs will have access to private data.
Look Mark, I assume you're a photographer. Spin your favorite images
through ACR and through DPP or Nikon View or C1 or Leaf or whatever
the converter is of the company who made your camera and compare the
color for yourself. You don't need Ulli or me for that.
Edmund
On Tue, Sep 9, 2008 at 4:32 PM, MARK SEGAL <email@hidden> wrote:
> Edmund,
>
> re (1) Agreed, it's not a dumb argument - even credible. But its
> implications are not necessarily conclusive from that argument alone. If
> they were conclusive, it would imply that there is sufficient performance
> variability between samples of the same camera model (even higher end
> prosumer and pro models) to defy any third-party (non camera manufacturer)
> attempt at creating a generically accurate profile - or let's say as
> accurate as the state of the profiling science allows them to be.
>
> Re (2) I have no reason to doubt your relative satisfaction with the
> procedures you describe, but what I was interested in exploring is whether
> you found DPP to be more "accurate" out of the box compared with LR2 using a
> standard test image such as a CC24.
>
> Mark
>
> ----- Original Message ----
> From: edmund ronald <email@hidden>
> To: MARK SEGAL <email@hidden>
> Cc: Uli Zappe <email@hidden>; ColorSync Mailing
> <email@hidden>; Chris Cox <email@hidden>
> Sent: Tuesday, September 9, 2008 9:37:49 AM
> Subject: Re: Media Testing for maclife.de (was: Re: Colorsync-users Digest,
> Vol 5, Issue 290)
>
> Mark,
>
> Canon know what sensor they build into a given sample of a camera. If
> they have to change something during production eg. the sensor is
> from a different batch, they burn the information into the camera ROM,
> I believe. DPP has access to this private calibration information, and
> will extract it from the file when attempting to convert. Third party
> converters that do not have access to ROM calibration info will just
> treat every camera like the original samples which were initially
> tested to write the software.
>
> I have talked to camera guys -not Canon- about this, and they have
> confirmed that they establish calibrations for every batch of sensors,
> and this information goes into the files written by the camera.
>
> I don't thing it's a dumb argument to say that a camera manufacturer
> knows their own camera best.
>
> (2) Doesn't make much sense to me. I have some files in my archives
> where the old versions of ACR failed spectacularly, but the software
> has evolved a lot since CS2. My routine at that point was to use ACR
> for the big patches of hundreds of runway images, and DPP for the few
> images which went full page. The last two years I have used Leica and
> Phase much more than Canon, and so have had little reason to use ACR
> on my own work. C1 yields good color when used with Phase, again
> because it is the manufacturer's converter.
>
> Edmund
>
>
> On Tue, Sep 9, 2008 at 3:18 PM, MARK SEGAL <email@hidden> wrote:
>> Edmund,
>>
>> Two questions:
>>
>> (1) In what ways does Canon's private info from the ROM give their
>> profiles
>> a heads-up over the procedure Adobe employs for the construction of camera
>> profiles? Or put otherwise, what aspects of Adobe's procedures, not having
>> that information directly, necessarily yield a less accurate profile than
>> those Canon constructs using its "private info"?
>>
>> (2) Can you publish test results for a standard, well-known photograph (or
>> a
>> GMB 24-CC) showing in what respects an out-of-the-box rendition of the
>> same
>> file from DPP is superior to the same from LR2?
>>
>> Mark
>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden