Issue #0702/1 - Now that the proposals for a standard methodology for testing of consumables for colour printers is in its final stage, manufacturers are beginning to quote figures to the new standard. We take a look at why the standard was necessary and how it affects the figures we will see in the future.
Even though formal ratification has not occurred yet, printer manufacturers are beginning to quote colour consumable yield figures to the new ISO standard, in advance of ratification, because the actual methodologies have been agreed and will not now change.
First to be seen presenting this data are Epson and Hewlett-Packard. It is the inkjet printer sector that has been tackled first, probably because of the lower number of pages involved to determine yield for a cartridge set. The approaches taken by theses two are different though. While Epson appears to be issuing figures for each new printer as it is released, Hewlett-Packard has presented data for its entire range of current inkjet printers.
So, with this comprehensive data from Hewlett-Packard, for which there is also pre-ISO data available, we take a look at four levels of device and compare the quoted yields under the old 5% scheme and under the new ISO standard. We then evaluate what this means in terms of visible Cost of Printing.
Before doing that though, we should remind ourselves of why this is an issue and what the ISO standard aims to achieve.
Lack of standards on testing and quoting of consumables yield accounts for one of the most significant difficulties that users have faced when attempting to decide on new hard copy output devices – particularly if economy is near the top of their agenda.
What the lack of standards means in practical terms is that figures quoted by the printer manufacturers for number of pages per cartridge are based on different metrics. Figures from a manufacturer for a printer in its range are only comparable against each of the other models in its own range and not against models from other manufacturers.
It may seem to be a simple matter to determine how many pages a cartridge produces – on the face of it, Yes, but in reality No! Debate has raged in the industry for many years over what constitutes a typical page; how yield should be measured; how it should be calculated; and how it should be quoted.
Under the old 5% scheme, every manufacturer developed its own test pages and test methodology and also decided how it was going to quote the results. This led to great confusion, lack of clarity and difficulty in being confident about comparability.
There are three major reasons why manufacturer’s data has been incompatible and incomparable.
- Different test pages have been used with different page coverage
- Testing has not been undertaken to a uniform methodology
- Yields may be quoted in different ways
Test pages
While for the majority of manufacturers it is impossible to know what sort of page the printer has been tested with, some manufacturers (e.g. Epson and Hewlett-Packard) do publish a thumbnail image of the test pages used in the tests either on the data sheet for the model or on the company’s web site.This is certainly very helpful but, any difference in test page means that a difference in page yield – thus making the test results incompatible. The only compatibility is with other models from the same manufacturer
What we need to identify in yield testing is where the differences lie between products. This can only be achieved by applying a standard page to the test procedure. It is no good adjusting a test page to ensure a precise 5% printed coverage from each printer, as some laboratories have done, because this is, in itself, an indication that the printers are behaving differently.
We have to know how they are behaving differently. Put potatoes into a slicer and you get sliced potatoes out - put potatoes into a masher and you get mashed potatoes out! We don’t want ‘normalised results’, we want results from ‘standard’ testing – meaning that the input (the test page) has to be identical for every machine tested. It is no good putting mashed potato into a slicer and expecting to get sliced potatoes out – it has to be raw potato.
Ultimately, a standard test page allows us to discover how the printer driver and the device itself handle the data and place the image on the paper in terms of how many pages can be printed. In particular, fonts and lines may be printed at different weights by different printers. It is important that we allow this to happen.
Although the de facto standard for page coverage in the printer industry has always been 5%, the copier industry has quoted to 6% and the fax industry to 3% (Slerex Letter). Just to complicate the matter further, the printer industry historically has also used the Dr Grauert Letter (ISO 10561), often misquoting it as 5 percent when it is actually measured at approximately 3.2 percent cover. A number of manufacturers have quoted yield at 4 percent and even at 2 percent (for IBM IPDS protocol).
Dr Grauert letterIn a study undertaken by Oki some years ago, to find out what the average office fax page coverage is, the resulting figure was 7.5 percent.
Not representative - no standard!!
In point of fact, it has been noted that a recent Epson inkjet machine (Stylus Photo 1290S) quoted black ink yield at a page coverage of 3.5%, meaning that the apparent yield, at first glance, appeared to be 66% higher than reality.
This is exactly the type of discrepancy that ISO 24712 is targeted at erasing.
Testing methodology
- How can two experts each taste a different wine, give their verdict and expect the public to know which is the better wine?
- How can two experts each view a different piece of art, give their verdict and expect the public to know which is the more pleasing masterpiece?
- How can two experts each test drive a different car, give their verdict and expect the public to know which is the better, safer and higher performance car that is the most fun to drive?
They cannot - tastes differ, palates differ, opinions differ and so different values are applied when making the judgements.
In order to achieve the most reliable judgment in these instances, a panel of judges is needed who will all inspect the same item and offer individual verdicts so that an overall result can be determined using an average of those individual verdicts.
Even this is not entirely foolproof. In the plethora of talent TV shows currently being aired (X-Factor, American Idol, Pop Idol, Strictly Come Dancing, Dancing with the Stars, Dancing on Ice, etc.) there are wildly differing opinions and verdicts, frequently causing uproar amongst the audience or competitors.
With regards to printers, each manufacturer has different procedures and methodologies by which toner yield is measured during and after the product development process and we can easily see that the figures quoted by different manufacturers cannot possibly be compared with assurance. The different opinions and values mentioned above relate to the different methodologies and procedures used in the testing of print engines.
Don’t misunderstand this! Each manufacturer has always known exactly how many of its own test pages to expect from its own printers. Each prints enough pages to ensure that the device is reliable and ready for market that there is lots of cartridge yield data available to them.
But that is not the point. The issue is that Hewlett-Packard is testing in one way while Xerox is testing in another, Brother in another and Kyocera Mita yet another – each methodology is different, with different test pages and different methods and practices.
While, technically, theory would say that each and every piece of paper contains exactly the same amount of toner or ink because it has been measured at 5 percent cover, this just is not the case. In fact, each and every piece of paper probably contains a different amount of ink or toner. Apparently the letter H draws more toner to it than other letters because its shape means that it commands a greater electrical charge around it when it is written to the photoconductor drum.
More significant than this though, is the fact that pages printed on one unit from the factory may be different from pages printed on another unit of the same model from the same factory of the same manufacturer. And, even more wild, pages printed using one toner cartridge / inkjet print head may be different from those printed on another identical unit, despite being used within the same printer unit.
Experience at CharisCo Printer Labs actually shows that even within the same toner cartridge on the same unit of the same printer model on the same day, toner consumption per 1,000 sheets printed can vary by more than 70 percent!
The most importance factor is that every printer, copier, fax or multifunction device must be tested in exactly the same manner, using exactly the same test pages and performed using exactly the same methods and processes, regardless of which manufacturer the model is from.
This is where the ISO standards offer the customer base a better level of information than previously available.
Incompatible quoting of yield
Taken a step further, each manufacturer is calculating the average yield in a different way and also quoting the results in a different way. There may even be instances where the tests or results are deliberately designed or manipulated in order to gain some advantage from some specific feature or technicality of the device in question.It is always assumed that the yield of a toner or ink cartridge has been quoted as an average of the test sample. Unfortunately this may not be the case or our understanding of which type of average has been used may be incorrect (e.g. arithmetic mean or median). Further manipulation could occur by ignoring certain cartridges where the tested yield acts as an outrider to the main body of cartridges or, alternatively, a manufacturer could test more than the specified number of cartridges and ignore the least productive cartridges.
In most instances we are not told exactly how the yield figure is quoted.
Again, we might assume it to be an arithmetic mean, but it may be quoted as ‘at least … pages’ or even ‘up to … pages’. A further complication exists if a manufacturer applies a confidence factor – e.g. 95% of cartridges will yield at least 10,000 pages.
How can we compare this with an arithmetic mean of 10,000 pages or ‘cartridges will provide up to 10,000 pages’ with any confidence?
Overall then, we cannot be sure of the nature of the figure quoted unless the manufacturer specifically details this in the specification. Even if we knew exactly what the quotation meant, we would be unable to draw a true comparison without having access to all the data from the entire test sample.
What the industry has lacked is an independent body with the funds to be able to test hardware independently and present the customer base with real life, comparable data.
ISO benefits and downsides
That situation now changes to some degree. We have already seen the introduction of the ISO 19752 for testing and quoting of yield for mono printing on mono laser printers and now we see the beginnings of the introduction of the equivalent ISO standard for colour printing on colour printers (ISO 24712).It is easy to see how the ISO standards offer a tremendous improvement to the situation – standard test pages, standard test methodology and standard way of calculating and quoting the yield.
Is this the end of the story though? There are actually still some downsides to the overall situation as it affects the calculations that users will need to use in their purchasing decisions.
What this standard does not really tackle is the printing of mono pages on the colour devices.
Much is made of the desire to enable customers to utilise just the one printer, a colour one, in their office environments. To achieve this, customers will need to see that the cost of mono printing on a colour device is not significantly higher than the cost of mono printing on a dedicated mono device. It must encourage users to print their mono pages on the colour machine.
For testing of colour printing yield, the test suite approved by the ISO comprises a set of five pages – a mostly mono letter, two data presentation pages, a newsletter page and a block/line test pattern.
This suite is printed as a five-page document, printed as continuously as the capacity of the paper tray will allow. The yield of each cartridge is recorded as the total number of pages printed at the time the cartridge is exhausted.
Because the test suite contains only one letter page (which is mostly black) and one newsletter page, it is unlikely that the balance of colour to mono is very representative of actual use. Even if the balance is reasonably representative, it does not allow for any variation in the proportion of mono pages to colour pages printed so that users can obtain a personalised estimate.
Single ISO test sheetWhat we need is a yield figure for mono only printing and a separate yield for colour printing (including black). Then, and only then can a truly representative calculation be undertaken that allows for printing of mono pages independently of colour pages.
So, as it stands, while being a great improvement, the ISO standards do not specifically achieve the goal of providing users with best-possible data to encourage the desired switch to colour printers.
To achieve the goal, colour printers must be tested to both standards, individually, in order to allow customers to produce a really definitive estimation of Total Cost of Printing for both mono pages and colour pages and a combination of the two.
|
Hewlett-Packard Inkjet |
Ink |
Quoted Yield (pages) |
Increase
%
|
|
|---|---|---|---|---|
| 5% | ISO 24712 | |||
| No. 88 ink set | Black | 2,350 | 2,580 | 9.8% |
| Cyan | 1,200 | 1,840 | 53.3% | |
| Magenta | 1,200 | 2,190 | 82.5% | |
| Yellow | 1,200 | 1,650 | 37.5% | |
| No. 363 ink set | Black | 800 | 1,160 | 45.0% |
| Cyan | 350 | 430 | 22.9% | |
| Magenta | 350 | 380 | 8.6% | |
| Yellow | 490 | 530 | 8.2% | |
| Light Cyan |
Assumed un-used |
8,800 | - | |
| Light Magenta | 11,700 | - | ||
| No. 339/344 ink set | Black | 800 | 870 | 8.8% |
| Tri-Colour | 450 | 580 | 28.9% | |
| No. 21/22 ink set | Black | 150 | 190 | 26.7% |
| Tri-Colour | 138 | 165 | 19.6% | |
It seems that it will be at the discretion of individual manufacturers as to whether the two standards are applied or just the colour standard. TCPglobal would wish to encourage manufacturers to move in the direct of testing every colour printer to both ISO standards.
With the figures provided under the new standard for colour printing only, however, we are certainly much more able to make a good comparison between printers, even from different manufacturers. We can inject the same split between mono and colour printing that we have done under the original system to give a better estimate of Total Cost of Printing.
Comparing old with new
Moving on to how testing under the new ISO standard will affect the figures we see on specification sheets, what we are seeing is a significant increase in the number of pages being quoted in the yield figures.In the accompanying figures extracted from Hewlett-Packard’s yield data, we see that this increase is never below 8% and may be as much as 82%!
Applying these yields to CPP and Total Cost of Printing calculations, the effect this has is to decrease perceived cost – by 9-31% when considering nominal CPP of the cartridges themselves and by 13-26% for long-term CPP for mixed mono/colour usage (ink costs only, not including hardware purchase price).
| UK - Inkjet | Printing |
Nominal CPP (UK pence) |
Difference
%
|
|
|---|---|---|---|---|
| 5% | ISO 24712 | |||
| No. 88 ink set | Office mono | 0.77 | 0.70 | -8.9% |
| Office colour | 4.21 | 2.91 | -30.9% | |
| No. 363 ink set | Office mono | 2.23 | 1.54 | -31.0% |
| Office colour | 6.85 | 5.73 | -16.3% | |
| No. 339/344 ink set | Office mono | 2.23 | 2.05 | -8.0% |
| Office colour | 6.96 | 5.72 | -17.8% | |
| No. 21/22 ink set | Office mono | 6.66 | 5.26 | -21.1% |
| Office colour | 15.35 | 12.52 | -18.4% | |
Once the standard has thoroughly taken root, there will be no problem with the new level of figures. Where there will be short-term confusion is where an organisation is attempting to replace a fleet of printers and wants to run a comparison of costs between the outgoing fleet and the short-listed replacements.
| UK - Inkjet |
Mixed mono/colour CPP over 3 years Ink Only (UK pence) |
Page volume 70% mono pages 30% colour pages |
Difference
%
|
|
|---|---|---|---|---|
| 5% | ISO 24712 | Pages/month | ||
| No. 88 ink set | 1.78 | 1.31 | 1,000 | -26.4% |
| No. 363 ink set | 3.54 | 2.68 | 500 | -24.3% |
| No. 339/344 ink set | 3.60 | 3.15 | 500 | -12.6% |
| No. 21/22 ink set | 8.99 | 6.99 | 100 | -22.2% |
In these circumstances, an uninformed user may think that a replacement has been found that is up to 26% less costly than the machines being replaced when this may not be the case. Note that the biggest reductions tend to be on the colour inks – the area in which the manufacturers particularly want to push a lower perceived Cost of Printing in order to help persuade customers that colour printing is not as expensive as it is currently perceived to be.
~End~
