Web Analytics RiceHigh's Pentax Blog: The Maker of the "ADC" and "IPU" Chips used in K10D

Thursday, December 07, 2006

The Maker of the "ADC" and "IPU" Chips used in K10D

After I posted my last article of “The K10D Secrets (Part 2) - The Bridge: ADC” yesterday, various people have responded that they had the doubt if Pentax really used the NuCore devices in the K10D, or not.

As I have mentioned in my last article, this is an *officially unconfirmed* *rumour*. By the most primitive definition, a rumour is something that you can’t get it confirmed by formal means. However, I really can’t understand why while most people can accept for the also a *rumour* (by definition) that a Sony 10MP CCD is being used in the K10D, some particular people cannot accept for the possibility that the ADC and IPU chips are actually come from NuCore.. Just because the company are not as famous and large?

Actually, this topic has being discussed repeatedly for so many times since late August, at various Pentax forums on the net. To sum up those numerous discussions, the following are some relevant factual findings for digest, for anyone who are interested:-

1. Appearances of the Chips

First of all, you can have a look on the chip appearances for the photos posted at the Dpreview K10D link below:-

Well, considering the close relationship between Dpreview.com and any camera manufacturer, I consider the photos and the associate descriptions published on the page to be at least semi-official, frankly speaking.

Now that first look at the “ADC” shot posted at Dpreview:-

.. and then compare it against the photo of the NuCore NDX-2240 chips in their technical catalogue:-

Next, look at the “PRME” engine photo posted at Dpreview:-

.. and then compare it against the photo of the chip in the NuCore SiP-2290 catelogue:-

It should be noted that there are no silk-printed marks on all the chips in the Dpreview photos. As such, all the readers will have no idea on the codes and the types of the chips used as well as the make. Anyway, it is very unusual for those chips used in a product do not have any marks and numbers on them, unless just in case those are intentionally left out or wiped out for some reasons.

Just for your further reference, here is a photo showing what a “normal” electronic chip should look like on its package:-

At this point, I bet that you should at least have some brief ideas after comparing the pairs of the different photos. So, simply merely judge by yourself!

2. Information about the Chips, as described by Dpreview:-

22 bit Analog to Digital converter
If accurate this could potentially deliver a very fine number of gradations, especially useful in shadow areas. However don't forget that this will always then be downsampled to eight bits per channel for JPEG conversion or stored as 12 bits in RAW. How relevant this '22 bit ADC' is to the final image quality is yet to be proven. ” – Phil Askey (owner of Dpreview.com)

Well, this is exactly the same design and features as the NuCore NDX-2240 12-bit model. Furthermore, as what the official K10D manual tells, the ADC of K10D is in 22-bit but RAW image data is only in 12 bits. But the crucial point is that it should be noted that a 22-bit ADC is not an ordinary thing. So, it may not be a coincidence.

'PRIME' Image Processor
Pentax has followed Canon and Sony in giving a name (a branding) to their image processor. For the K10D this unit is called 'PRIME', that said they aren't giving much away other than to state that it is built on 90nm production process and supports DDR2 (800 MB/sec) RAM which the K10D utilizes (and according to Pentax is the only SLR to do so). ” – Phil Askey

Again, in the NuCore technical catalogue (in the above provided link), it can be seen that their SiP-2290 particularly uses DDR2 RAM. The utilisation of DDR2 RAM is an unique feature of the K10D, as claimed by Pentax, according to the what Dpreview report, as quoted in the above.

3. BTW, here is another quote on "the official “take” on the 22-bit A/D Convertor":-


It would be very interesting to note the used terms of “analog signal re-conditioning” and “oversampling” by Pentax Japan (for what the post says) are actually two of the key features highlighted by NuCore in their (technical) marketing information at their website and in their catalogues! What a coincidence!

BTW, here is some food for thought:- Does the original device maker really matter? I think in the end, only the end results and the actual performance of the camera do really matter!

So, if good chips, which have all those powerful features (I mean the NuCore ones), are actually used in the K10D, it should be a very good news afterall. However, sometimes I still wonder why there are still some people (whom are usually those die-hard blind brand loyalists, as usual) “bashed” me over and over again at the net for the simple facts and technical knowledge which I intended to share, here is a latest example.

My humble opinion is that even though currently there are many "outside" evidences show that the K10D is quite likely having the NuCore chips in it, the OEM brand indeed doesn't really matter in the end! However, it would be always interesting to know more about the NuCore devices, under this scenario, at least for me or probably for everyone who are interested about the K10.

All in all, if the K10D really performs, I would regard Pentax’s engineers have done a great job and they have successfully completed a beautiful system integration, for all those different hardware devices from various manufacturers to software (firmware) programming (Well, Pentax do market the SilkyPix which is an OEM software, in contrast).